nlevel


SYNOPSIS

nlevel  [level=<2-4 or ALL; defaults to ALL>]  \
    [field_list=<comma/hyphen separated list | ALL; defaults to ALL>]


DESCRIPTION

The nlevel keyword is used to select n-level variables, that is variables assuming only n different values across the various objects of the dataset with an uneven distribution among them. In particular, 2-level variables are those variables assuming only two different values across the dataset, one of which is found only in 2, 3 or 4 objects. 3-level variables are those variables assuming only three different values across the dataset, two of which appear only in 1 or 2 objects. 4-level variables are those variables assuming only four different values across the dataset, one of which appears only in one object, while the other three may appear in up to 2 objects. According to Baroni et al. [1], it is advisable to remove such variables having a highly skewed distribution of values before building the PLS model since they could put a heavy bias on the latter.

EXAMPLES

# the following command identifies 2, 3, 4-level variables on all fields; once identified, they are removed by means of the remove_x_vars keyword
nlevel
remove_x_vars  type=NLEVEL


REFERENCES

  1. Baroni, M.; Costantino, G.; Cruciani, G.; Riganelli, D.; Valigi, R.; Clementi, S. Quant. Struct-Act. Relat. 1993, 12, 9-20.   DOI

Sitemap
Print version
Contact
Mailing list


Last update:
May 31. 2015 20:39:42

Powered by
CMSimple - CMSimple-Styles


Get Open3DGRID at SourceForge.net. Fast, secure and Free Open Source software downloads



Would you like to align your
dataset? Try Open3DALIGN
Just wish to compute a MIF?
Try Open3DGRID